Feasibility of measuring walking speed in a diabetes clinic in primary care

Abstract Background Usual walking speed may predict cardiovascular disease in type 2 diabetes, but it is not routinely assessed in UK primary care. Objective To examine the feasibility of three methods of assessing walking speed in patients with diabetes in general practice: timed walking, accelerometry, and treadmill walking. Methods We approached consecutive mobile patients attending a general practice diabetes clinic. Consenting participants performed a timed 28-m walk (4 × 7 m) at their usual pace in the surgery corridor while wearing an accelerometer, followed by a 3-minute treadmill walk. Measures were repeated after 6 months. Differences in walking speeds between methods and between timepoints were assessed by Bland–Altman limits of agreement. Results Recruitment rate was 91% (30/33). Mean age of participants was 59.6 years (range 37–75), mean body mass index was 30.6 kg/m2 (21.0–46.4), and 17% had a history of cardiovascular disease. Average duration of diabetes was 10 years (range 1–40). Twenty-five participants (83%) repeated assessments at 6-month follow-up. Average (SD) walking speed on the timed walk was similar to both accelerometer and treadmill: 1.12 (0.13), 1.15 (0.16), and 1.01 (0.39) m/s, respectively. Bland–Altman analysis showed good agreement for repeat tests and, on average, between methods. Limits of agreement between timed walk and treadmill were wide. Feedback from seven participants and three practice staff indicated acceptability of all tests, but treadmill testing took longer and least feasible. Conclusions Timed walk was the most feasible and acceptable to patients and staff. Results were similar on average to accelerometry and treadmill.